The EDA of Crypto: Why Undervalued Enablers Are the Next Narrative Frontier

In the AI boom, the CEO of Cadence Design Systems publicly argued that his company is undervalued. The semiconductor industry largely dismissed it as executive posturing. But for those who track narrative cycles, the statement was a signal. Cadence, the EDA (Electronic Design Automation) giant, sits at the base of every AI chip design. It is the quintessential "pick-and-shovel" vendor in the gold rush of AI infrastructure. Yet, its market cap remains a fraction of the AI chipmakers it enables. The same structural undervaluation is now emerging in blockchain infrastructure. The protocols that provide the foundational compute, data, and coordination layers for decentralized AI are being priced as commodities, when in fact they are becoming the EDA of the crypto-AI stack. History repeats, but the narrative layer shifts. In 2021, the crypto market chased application-layer tokens—DeFi, NFTs, gaming. In 2024, the narrative shifted to AI agents and autonomous systems. But the market is still looking at the end products, not the invisible rails. Cadence’s story is a mirror. Over the past decade, its revenue grew from $2 billion to over $4.6 billion, yet its price-to-sales ratio hovered around 10-12x, while Nvidia’s soared past 30x. The reason is simple: the market sees EDA as a mature software tool, not as a leveraged bet on AI chip complexity. The same blindness is happening in crypto. Layer-1s like Ethereum, data availability layers like Celestia, and compute networks like Akash are the "EDA" of the crypto-AI stack. They are the critical enablers, yet their valuations ignore the "value concentration" effect—the fact that each dollar of infrastructure revenue supports hundreds of dollars of downstream application value. Every chart is a frozen moment of human emotion. The Cadence undervaluation claim is rooted in three hidden mechanisms that apply directly to crypto infrastructure. First, the "tool tax" effect: as AI chips become more complex, the cost of EDA tools per chip rises exponentially. From 4nm to 2nm, design costs balloon from $200 million to $500 million, and EDA’s share stays at 25-30%. This means Cadence’s revenue per chip is growing faster than the chip market itself. In crypto, the same logic applies to gas fees and data availability costs. As AI agents execute more on-chain logic, the demand for block space and compute increases non-linearly. Protocols like Celestia or EigenLayer capture a fraction of that value, but the growth rate of their "unit economics" per agent is accelerating. Second, the "platform transition" effect: Cadence is moving from software licensing to a platform model (cloud-based, AI-embedded, subscription). This expands its addressable market from $100 billion (pure EDA) to $300 billion (system-level design). Similarly, blockchain infrastructure is evolving from simple settlement layers to composable platforms for AI agents. Ethereum’s rollup-centric roadmap, for instance, turns it into a unified settlement layer for thousands of autonomous agents, each paying fees for security and finality. The market still prices Ethereum as a payment network, not as an operating system for AI. Third, the "geopolitical hedge" effect: Cadence’s strategic importance in export controls has made it a national security asset, creating a floor for its valuation. In crypto, the push for "digital sovereignty" is pushing governments to adopt blockchain for identity, supply chains, and even AI auditing. Protocols that offer verifiable compute or decentralized identity become infrastructure assets, not just speculative tokens. The core of the undervaluation lies in the narrative lag. The market is still using old frameworks. For Cadence, analysts used traditional software metrics (P/E, revenue growth) and ignored the leverage effect. The same is happening in crypto. Let’s quantify the gap. Ethereum’s annualized fee revenue is roughly $2.5 billion (post-Merge, post-L2 scaling). Its market cap is around $300 billion. That gives a P/E of 120x—high by traditional standards. But if we apply the EDA leverage ratio (each dollar of EDA revenue supports $200-300 of semiconductor value), then Ethereum’s $2.5 billion in fees supports an estimated $500-750 billion in downstream application value (DeFi, NFTs, gaming, and now AI agents). That implies a leverage ratio of 200-300x, similar to Cadence. Yet the market does not price Ethereum as a leveraged enabler; it prices it as a volatile asset. The blind spot is systematic. For blockchain-native AI projects like Bittensor (TAO) or Render Network (RNDR), the situation is even starker. Bittensor’s subnet infrastructure allows AI models to be trained and served on a decentralized network, but its market cap of $3 billion is tiny compared to the $200 billion+ in centralized AI capex. The market perceives Bittensor as a niche experiment, not as the "EDA" for decentralized AI. Yet the same logic holds: every AI model deployed on Bittensor requires subnet compute, generating fees for the network. The leverage ratio is enormous, but the narrative hasn’t caught up. Clarity emerges only after the noise subsides. The contrarian angle is that these infrastructure tokens are actually overvalued because of hype cycles. But the data suggests otherwise. The hype is centered on applications—AI agents, memecoins, NFTs—while the infrastructure tokens trade at discounts to their fundamental value. Consider the case of Celestia. Its data availability layer is designed to scale rollups, which are the primary execution environment for AI agents. In 2025, the number of rollups using Celestia grew 500%, yet TIA's price lagged. The market is pricing it as a commodity blockchain, ignoring the fact that its revenue is tied to the growth of the entire rollup ecosystem. The same pattern is visible in the modular blockchain thesis. The infrastructure layers (data availability, consensus, execution, settlement) are the EDA tools of the crypto world. They are the "boring" middle layers that enable the exciting applications. But the market rewards the applications, not the enablers. This is a narrative arbitrage opportunity. The bear market of 2023-2024 washed out the hype, revealing which protocols have real usage. The surviving infrastructure tokens—Ethereum, Celestia, EigenLayer, Akash, Bittensor—have demonstrated product-market fit. Yet their valuations are still depressed relative to the potential of the AI-crypto convergence. Takeaway: The next narrative shift will be from "which AI agent wins" to "which infrastructure enables them all." The Cadence story is a warning to crypto investors: don't ignore the enablers. When the CEO of a critical infrastructure company says it's undervalued, listen. The same logic applies to the protocols that are the EDA of the crypto-AI stack. The market will eventually realize that the value is concentrated in the infrastructure, not the applications. The code is permanent; the meaning is fluid. Right now, the meaning of "blockchain infrastructure" is undergoing a revaluation. The investors who see the parallel will capture the next narrative wave.